steam conference on vr
DESCRIPTION
What VR Should Be In 10 Years.TRANSCRIPT
MICHAEL ABRASH VALVE WHAT VR COULD, SHOULD, AND ALMOST
CERTAINLY WILL BE WITHIN TWO YEARS
The near future of virtual reality
• VR is likely to be here sooner than you think
• We’ve learned this from our R&D
• We’ve built a hardware prototype that creates presence
The near future of virtual reality
• Presence is VR magic
• We think people are going to want presence badly
The near future of virtual reality
• The same hardware reduces motion sickness
• This hardware makes it possible to create incredible VR
experiences that won’t make people sick
• We believe that a consumer version is feasible within
two years
Who might ship this? • Oculus is the obvious candidate
• The DK1 is a good first step, but not good enough
• Crystal Cove is a big step in the right direction
Resolution, latency, persistence, translation
Low persistence driven by Valve prototyping
Valve and Oculus collaborated on tracking
• We’re continuing to work with Oculus to drive PC VR
forward
Implications of presence
• VR could evolve into a major platform
• Presence could tip the balance of the entire
industry toward computer entertainment
• Presence requires a head-mounted display
connected to lots of local compute power
• The PC is going to be the best place for VR
Decide for yourself
• Try it for yourself and decide what you think the
future looks like
WHAT DOES IT TAKE TO MAKE VR
COMPELLING?
Murray, S.O., Boyaci, H., and Kersten, D.
(2006). The representation of perceived
angular size in human primary visual cortex.
Nature Neuroscience. 9, 429-434.
doi:10.1038/nn1641
http://www.bilkent.edu.tr/~hboyaci/Vision/SizeAppletLarge.html
Fang, F., Boyaci, H., Kersten, D., and Murray,
S.O. (2008). Attention-dependent
representation of a size illusion in human V1.
Current Biology, 18 (21), 1707-1712.
doi:10.1016/j.cub.2008.09.025
http://www.echalk.co.uk
http://www.echalk.co.uk
Presence
• The sense that you are really someplace else
• More than just immersion
• Can only understand by experiencing it
Presence
• Varies from person to person
• Has gotten stronger as technology has improved
Presence
• It’s why we’re so excited about VR
• Powerful at a visceral level
• Unique to VR
• Likely to be the key to VR’s success
All of the following are needed • A wide field of view
• Adequate resolution
• Low pixel persistence
• A high enough refresh rate
• Global display
• Optics
• Optical calibration
• Rock-solid tracking
• Low latency
A wide field of view
• Immersion
• Peripheral cues
• At least 80 degrees
• More is better
Adequate resolution
• Problematic for VR due to wide field of view
• One-seventh the pixel density per degree of a
wide-screen TV
• 1080p works
• More is better
Low pixel persistence
• Necessary to avoid blurring with eye motion
• No more than 3 ms
Smear from persistence
A high enough refresh rate
• Necessary to avoid flicker with low persistence
• 95 Hz seems to be sufficient
• Somewhat less than 95 Hz may be adequate
Global display
• Illuminates all pixels simultaneously
• Avoids motion-induced compression, stretching,
and skewing
• Rolling display may work, but will have failure
cases without low-latency eye tracking
Optics
• Only one or at most two lenses per eye
• Not enough to allow simultaneous control of
Focal length
Viewing distance
Size
Distortion
Various forms of aberration
Optics
• Search for best tradeoffs in a huge space
• Evaluation has to be done by a human
Optical calibration
• The human visual system is amazingly sensitive to
deviations
• Many problems are impossible to identify until
calibration is nailed
Tracking
• Must support translation (position in x, y, and z) as
well as orientation
• Position with at least millimeter accuracy
• Orientation with at least quarter-degree accuracy
• Volume at least 1.5 meters on a side
Low latency
• 20 ms motion-to-last-photon works
• 25 ms may be good enough
Presence is emergent
• All of these factors have to be good enough
• No one factor is enough by itself
SO WHEN IS PRESENCE COMING?
Feasible 2015 consumer HMD • 20 ms motion-to-last-photon latency
• 3 ms pixel persistence
• 95 Hz refresh
• 110-degree FOV
• 1K x 1K resolution per eye
• High-quality, well-calibrated optics
• Tracking
millimeter-accurate resolution translation
quarter-degree-accurate rotation
volume of roughly 2 meters cubed
Presence in 2015
• We’ve built prototypes to this spec
• We believe the technology is transferable to
consumer-priced head-mounted displays
• We’ll share what we’ve found with PC companies
that want to develop VR hardware
There’s a lot left to be done
• Improve every key element
Up to 100X resolution would help
Optics are far from optimal
Head tracking isn’t fully solved
Eye tracking is far from solved
• Solve per-user lens positioning (IPD, eye relief)
• Get rid of the tether
• Get a display manufacturer to make VR-optimized panels
There’s a lot left to be done
• 3D audio
• Haptics
• Body tracking
• Input
• Especially the interaction between input and
game design in VR
VR software • Movement
• Gameplay
• Multiplayer
Could be the best online social experience yet
• Content
This is where you come in • Could be one of the biggest platform shifts ever
• Platform shifts create opportunities
• Great VR requires custom experiences
• Someone is going to write the DOOM of VR
The PC will be the hotbed for VR
• PC – Linux, Windows, OSX - will be the VR epicenter
Rapid evolution
• Already happening with the Rift
Many developers and IHVs, with complete
freedom to innovate
Most powerful platform, and getting more so
Valve will continue to help drive VR on the PC &
Steam forward
Takeaway
• A great VR system at a consumer price in 2015 is
primed to happen
• It will happen, in 2015 or soon after
Thank you
• Find more information and join the conversation at
http://blogs.valvesoftware.com/abrash/